GNU/Linux Is Not That Other OS, Again…

I came across a post in a forum challenging GNU/Linux experts to find a file-manager in GNU/Linux that would allow the authour to use GNU/Linux exactly the way he uses that other OS:“1. Search selected network drives / partitions / directories for files by name using wildcards
2. Place the results in a window so that one or more can be selected for cut, copy, edit, rename and delete
3. When files are copied or moved, it is essential that the "date modified" property is preserved.”

Of course, he wants a GNU/Linux file-manager to do things exactly the way that other OS does, which is silly. The bizarre search path string he enters is definitely not something anyone would likely ever type. It’s more likely a constant for long periods of time, exactly like a file-system… In GNU/Linux one would mount all the appropriate directories and file-systems in a tree-structure, index them with a search engine and browse them neatly with the typical GNU/Linux file-manager. I suppose this fellow could expect an OS to hunt all over the tree structure for every search like XP did in the ancient times but it’s so much more intelligent to index the thing periodically and perhaps just reindex the parts liable to frequent changes.

I have a similar structure in my home, albeit with fewer computers, but the little woman can roam with a notebook or tablet or sit at her desk and search everything. I recommend recoll for a desktop search engine. It’s easy to schedule indexing and it’s no problem to select certain file-types or locations to index. Does recoll’s result window open in the file-manager? Nope, but I can examine the files and select paths to place in the file-manager as I wish. The result window is an html page that takes you to the browser or other application assigned by mime-types. You can format the page to just hold URIs for the search-result files and pass them to the file-manager or you can strip off the filenames and pass the directories to thunar.

Recoll uses Xapian as a search engine/database and it’s extremely scalable.

That may not be exactly the way M$ does it but it’s much more configurable by a long shot and infinitely flexible. I can use a different file-manager if I want and I can tweak the source code.
recoll -q "ext:txt mips eating"

This just outputs the basic file URIs and passes the list of directories to the file-manager:dirname `recoll -b -t -q "ext:txt kill netscape"`|uniq|thunar `cat -`

See Forums • View topic – Why I can't leave Windows XP for Linux.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , , , . Bookmark the permalink.

6 Responses to GNU/Linux Is Not That Other OS, Again…

  1. matchrocket wrote, “It’s not fast like a pre-indexed search but that’s the trade-off I’m willing to make.”

    That may be just fine on a single-user system with just a few files but these days, ordinary folk have hundreds of thousands of personal files and multiple storage devices local and over the network. Indexing is essential IMHO. For multiusers, I like Swish-e because it can do local file-system and the Internet or LAN all at once. If you have enough RAM, you can index anything. I took thousands of files to the North and used Swish-e to provide nice search for whole schools. It’s like a library that dwarfed their dead-tree kind but with few seconds per search. It’s a bit fussy to configure but scales very well if you put it on a web-server. I’ve used it for huge repositories of HTML, text and recipes. It’s a huge asset at home, in classrooms and in schools. For a single PC or with mounted networked file-systems Recoll is very smooth.

  2. matchrocket says:

    I use Dolphin in KDE. It has a Find function that can find a file name or a string in the contents. I can specify Documents, Images, Audio or Video files, and a range of times. It’s not fast like a pre-indexed search but that’s the trade-off I’m willing to make.

    I don’t like indexed search functions. They bug me when they’re indexing. I don’t like my computer doing things on it’s own. I don’t trust it. Just call me paranoid.

  3. dougman says:

    Reviewing pricing using M$:
    (7) $3500 server licenses
    (700) $50 CALs
    NET: $60K USD

    Obscene pricing by M$!

    A single server with failover and off site redundancy could be built for 1/3 the cost easy.


  4. dougman wrote, “Typically i just use” locate filename” it finds it in under 1-2 seconds flat.
    Dude running 9 computers for a make-shift NAS for a total storage of around 24TB??”

    “Locate” finds files by file/pathname, not contents. That’s useful of course if names are memorable/indicative but this guy used the alphabet… It certainly would be more efficient to use fewer boxes. In the past, I’ve built ATX systems with 10 drive bays in a standard box. You could put 30TB on a system like that but of course that’s built from scratch, not incrementally. I once worked in a place that had 7 of M$’s servers, mostly because of turf-wars between departments and the fact that they grew storage by adding servers, not drives. Some people just love to send M$ money. That’s the same place that figured out they weren’t properly licensed and forked over tens of $thousands to make things right. Figure it out… 7 server licensing fees plus 100 seats of “pro” this and that plus 700 cals… The place was an advertisement for Free Software. I did get to install one GNU/Linux server in a virtual machine and a lab of clients was brought back from the dead by converting my old motherboard for Beast to a terminal server. It’s just amazing what some people will endure to support M$.

  5. dougman says:

    He also should use Google online storage in case of a house fire, 20TB is like $200/yr.

  6. dougman says:

    Typically i just use” locate filename” it finds it in under 1-2 seconds flat.

    Dude running 9 computers for a make-shift NAS for a total storage of around 24TB?? LOL… I have a Slackware built 4U in my downstairs rack serving up twice that and it is in one box. He needs to build one server and get rid of all the PC’s

    Wow, all that information stored on vulnerable Win-Dohs thats crazy.

    If he states he cannot leave Win-Doh’s let him stay there. He will learn from experience that his solution is persona non-grata in the long term.

Leave a Reply